[From Bruce Gregory (2004.0629.0727)]
I believe that some time ago I posted the following little story. I was driving to work and passed a dry cleaning store. I remembered that I had left a suit there last week and turned into the parking lot to retrieve the suit.
Now it is not difficult to imagine a control theory model that would account for this behavior. In one such model I am controlling the perception, "If you are near the dry cleaners and more five days have passed since you dropped something off, stop and retrieve it. I can imagine this program constantly running in the background. Normally there is no error generated by not thinking about the dry cleaning or by not stopping by to pick it up. However, when I see the dry cleaner, error would be generated if I did not turn into the parking lot. So I do.
Does this story explain my behavior? Is anything missing? Should I be satisfied with a model of my behavior that consists of a very large but finite set of control circuits based on similar if-then perceptions?
Bruce Gregory
There seems to be no shortage of people who value certainty above truth.
[From Hank Folson (2004.0629.0900)]
From Bruce Gregory (2004.0629.0727)
Does this story explain my behavior? Is anything missing?
What can appear as an "if-then" situation is probably handled in a complex living control system by a gradual changing of the size of the error signal. The "pick up dry cleaned suit" goal has almost no error signal on the day you dropped the suit off. Each day that passes, the error signal increases. By day 5, the error signal becomes large enough that, barring some conflicting higher priority goal, you will pull into the dry cleaner's parking lot.
Sincerely,
Hank Folson
[From Bruce Gregory (2004.0629.1406)]
Hank Folson (2004.0629.0900)
from Bruce Gregory (2004.0629.0727)
Does this story explain my behavior? Is anything missing?
What can appear as an "if-then" situation is probably handled in a complex living control system by a gradual changing of the size of the error signal. The "pick up dry cleaned suit" goal has almost no error signal on the day you dropped the suit off. Each day that passes, the error signal increases. By day 5, the error signal becomes large enough that, barring some conflicting higher priority goal, you will pull into the dry cleaner's parking lot.
Interesting idea, but I have no idea how to model it. As I understand it, the error is the difference between the perceptual signal and the reference level for that perception. (Bill will correct me if I have this wrong.) Why does it grow with time? it doesn't seem that the perceptual signal changes until I pick up the suit. Does the reference level change? Is it being changed by a higher level system? What might that system be?
Bruce Gregory
There seems to be no shortage of people who value certainty above truth.